Real-Time Gaze Estimation Using Monocular Vision
Improving the accuracy of gaze estimation and the tolerance of head motion is a common task in the field of gaze estimation. The core problem of gaze estimation is how to accurately build up the mapping relationship between image features and gaze position. To this end, we propose a method to reconstruct input features in the optimized subset as the key to our solution. The HOG feature is considered as the input feature. First, we found the closest calibration point to gaze position and constituted the optimized subset. Then, we get a set of weights that can linear reconstruct test samples in the optimized subset. And this set of weights is used to express the mapping relationship. At last, a linear equation is fitted to solve head motion problem. In this paper, the experiment results demonstrate that our system can achieve high accuracy gaze estimation with one camera.
KeywordsGaze estimation Feature reconstruction Head move compensation Optimized subset
This research was funded by National science and technology support plan “User evaluation technology and standard research of display and control interface ergonomics”(2014BAK01B04).
- 6.Timm, F., Barth, E.: Accurate eye centre localisation by means of gradients. In: VISAPP, pp. 125–130 (2011)Google Scholar
- 7.Coutinho, F.L., Morimoto, C.H.: Improving head movement tolerance of cross-ratio based eye trackers. Int. J. Comput. Vis. 101(3), 1–23 (2012)Google Scholar
- 8.Sugano, Y., Matsushita, Y., Sato, Y.: Appearance-based gaze estimation using visual saliency. IEEE Trans. Pattern Anal. Mach. Intell. PP(99), 1 (2012)Google Scholar
- 10.Cerrolaza, J., Villanueva, A., Cabeza, R.: Taxonomic study of polynomial regressions applied to the calibration of video-oculographic systems. In: Proceedings of the Symposium on Eye Tracking Research & Applications, pp. 259–266. ACM (2008)Google Scholar
- 12.Xu, L.Q., Machin, D., Sheppard, P.: A novel approach to realtime non-intrusive gaze finding. In: BMVC, pp. 428–437 (1998)Google Scholar
- 13.Tan, K., Kriegman, D., Ahuja, N.: Appearance-based eye gaze estimation. In: WACV, pp. 191–195 (2002)Google Scholar
- 14.Ren, S., Cao, X., Wei, Y., et al.: Face alignment at 3000 FPS via regressing local binary features. In: 2014 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 1685–1692. IEEE (2014)Google Scholar
- 15.Wright, J., Ganesh, A., Zhou, Z., et al.: Demo: Robust face recognition via sparse representation. In: IEEE International Conference on Automatic Face and Gesture Recognition, Fg 08, pp. 1–2 (2008)Google Scholar